1013

Introduction: Acute kidney injury is a common complication in patients treated with extracorporeal membrane oxygenation (ECMO). Continuous renal replacement therapy (CRRT) can help to optimize fluid status and reduce the inflammatory response associated with ECMO. However, published data are derived primarily from children without primary cardiac disease. Few data are available on the impact of CRRT on the survival of critically ill patients receiving ECMO support. Methods: A retrospective analysis of our institutional ECMO database (n=144) from November 2008 to June 2013 was performed. Patients with ICU survival < 24 hours were excluded. We collected demographics, co-morbidities as well as concomitant therapy in all patients treated during the study period. Chronic renal failure was defined as a serum creatinine ≥ 2.0 mg/dL before ICU/hospital admission. Severe bleeding was defined as the need for ≥ 4 red blood cells packs transfused over 24 hours. Patients treated with CRRT were compared to those not treated, regardless of their renal function. Mortality was assessed at ICU discharge. Results: Of 102 ECMO adult patients, 52 (51%) received CRRT. Three of them were treated twice during the ICU stay. The time from ECMO insertion to CRRT initiation was 2 [0-3] days. Patients on CRRT had similar age (53 [40-59] vs. 44 [32-56], p= 0.12) and similar rate of male patients (27/25 vs. 32/50, p= 0.21), diabetes (12/52 vs. 12/50, p= 0.91), chronic renal failure (5/52 vs. 1/50, p=0.20) and medical admission (48/52 vs. 40/50, p=0.07). Only the use of immunosuppressive agents was higher in patients with CRRT when compared to others (22/52 vs.7/50, p= 0.001), The proportion of veno-arterial ECMO was similar between the two groups (29/52 vs. 22/50, p= 0.23), as the APACHE II score on ICU admission (24 [18-28] vs. 22 [19-24], p=0.11), the duration of ECMO (6 [4-10] vs. 8 [4-9], p=0.68), the occurrence of infection (42/52 vs. 40/50 p=1.0) and severe bleeding (28/52 vs. 20/50, p= 0.17). Mortality was also similar between patients on CRRT (31/52) when compared to others (23/50, p=0.23). Conclusions: The use of CRRT was not associated with an increased mortality in an adult population of patients treated with ECMO with similar characteristics.